Mô tả

Gaali in Hindi means Slang which is Toxicity in today’s jargon

Stop toxicity and form slang – Gaali effectively eliminates harmful and unwanted comments, preemptively preventing any potential damage.

Ảnh màn hình

  • This is how the comment list in the backend looks with details of the toxicity parameters

Hỏi đáp

After posting a comment, why does it still say it’s in moderation ? Isn’t this supposed to be automated ?

Yes, it’s automated and there is no need for the wordpress admin to approve a good comment.
But to validate a comment for foul language and slangs, which is called a toxicity check, it takes time, like many many seconds, to detect it.

Why does it take so much time for a toxicity check ?

The toxicity check connects to TensorFlow, here using TensorFlowJS, which is a Google’s product and the data sets used are on Google’s servers.

So this connects to some 3rd party backend-service ?

Yes, it connects to an API at toxicity.co.in which is again, my (free) service which runs on Deno at Deno Deploy – deno.com which in turn goes to Google for returning the results of toxicity. Sometimes it takes a minute or more to return the result and at times, it even times out. If it times out the admin has to manually validate the comment in wp-admin.

Will this plugin remain free forever ?

I want it to remain free forever and open-sourced.
There is so much of free stuff on the internet and even open-sourced.
I want to be part of the community and don’t intend on commercializing it.
But if toxicity service at my deno.com account take a hit on many requests, I will then have to shell out minimum $10/mo ($2 per million requests per month, $0.30/GiB outbound data transfer)
Even then that’s okay and I can fund that minimum amount, but if it goes beyong $10 a month, then I will seek funding of some sort.

Đánh giá

There are no reviews for this plugin.

Người đóng góp & Lập trình viên

“Gaali” là mã nguồn mở. Những người sau đã đóng góp vào plugin này.

Những người đóng góp

Nhật ký thay đổi

0.5

  • Just started, will update when I come across issues.
  1. Comment check runs in the background
  2. Neat table in comment’s admin showcasing 6 toxic parameters and a master parameter Toxicity that will decide if the comment should be auto-approved or not
  3. API Key not required for now, but will incorporate this if there are many users of this plugin and the usage is high